| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, University of Miami School of Medicine, Jackson
Memorial Hospital, West Wing 279, 1611 N.W. 12th Ave., Miami, FL 33136.
2
A. G. Holley State Tuberculosis Hospital, 1199 W. Lantana Rd., Lantana, FL
33462.
Received October 19, 1998;
accepted after revision June 11, 1999.
Address correspondence to J. E. Fishman.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. A retrospective review of sequential chest radiographs was performed of 87 patients undergoing therapy for pulmonary tuberculosis: AIDS patients receiving antiretroviral therapy (n = 31), HIV-positive patients not receiving antiretroviral therapy (n = 26), and HIV-negative patients (n = 30). Pulmonary consolidations, thoracic lymphadenopathy, and pleural effusions were evaluated for worsening, stability, or improvement. Patients with concurrent pulmonary infections were excluded.
RESULTS. Transient worsening on radiography was observed in 14 (45%) of 31 AIDS patients receiving antiretroviral therapy, including seven patients (23%) who showed severe worsening. Of 56 patients in the other two groups, 11 (20%) showed worsening (p = 0.023), two of whom showed severe worsening (p = 0.009). Worsening was first noted between 1 and 5 weeks after initiation of antiretroviral therapy, with improvement occurring between 2 weeks and 3 months later. Four patients with severe worsening converted their tuberculin purified protein derivative responses from anergic to positive after antiretroviral treatment.
CONCLUSION. Transient worsening is frequently seen on chest radiography in AIDS patients with tuberculosis who subsequently undergo antiretroviral therapy. This phenomenon may be related to improved immune function.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Antituberculous therapy consisted of multidrug regimens according to the recommendations of the Centers for Disease Control and the American Thoracic Society [11, 12]. All patients were tested for HIV status at admission. For HIV-positive patients, CD4+ cell count was measured at the time of diagnosis, the day before antiretroviral therapy was begun, at subsequent 2-month intervals, and when worsening of signs or symptoms was observed. In most patients, multidrug antiretroviral therapy consisted of zidovudine (AZT), lamivudine (3TC), and either zalcitabine or didanosine. HIV-positive patients who were not receiving rifampin as part of antituberculous therapy received AZT, 3TC, and saquinavir (a protease inhibitor). All patients were known to have TB and had received some antituberculous therapy before admission. Because of the complexity of multidrug regimens for both TB and HIV, antiretroviral therapy was not started until patients had received at least 2 weeks (generally longer) of inpatient antituberculous therapy. Tuberculin purified protein derivative (PPD) (Connaught Laboratories, Swiftwater, PA) testing and anergy skin testing were performed on all patients at admission; if anergic, tests were performed every subsequent 2 weeks during hospitalization or until PPD conversion. Posteroanterior and lateral chest radiographs were obtained at admission and subsequent posteroanterior and lateral or anteroposterior radiographs were obtained at monthly intervals for the duration of the patient's hospitalization. Some patients exhibited clinical worsening during treatment, which was defined as new fever equal to or exceeding 101.5°F of at least a 1-week duration, or as worsened or new respiratory symptoms, cervical adenopathy, cutaneous tuberculous lesions, or ascites. Worsening prompted physical examination; cultures of sputum, blood, and urine; and additional examinations on chest radiography. Bronchoscopy, biopsy, or both were routinely performed if there were localized findings and physical examination and cultures did not explain worsening. Drug sensitivity was evaluated by performing monthly testing in patients who continued to produce sputum that was culture-positive for TB.
Of 98 patients admitted to the hospital during the study period (February 1996 through January 1997), three patients were omitted because their admissions were too short to generate a series of radiographs for review. Of the remaining patients, 40 had AIDS and 55 were HIV-negative. Seven of the AIDS patients were excluded because of nontolerance or refusal of antiretroviral therapy or because antiretroviral therapy was given before admission. The remaining 33 AIDS patients were termed group 1. Of the HIV-negative patients admitted during the study period, 31 were randomly selected from medical record numbers to obtain an approximately equal number of HIV-positive and HIV-negative patients for comparison; the HIV-negative patients were termed group 2. In designing the study, we sought to evaluate the possibility that worsening seen on chest radiographs might be caused by HIV positivity rather than antiretroviral therapy. Consequently, we formed a historical control group of all HIV-positive patients treated for TB at the hospital between January 1986 and December 1987 (n = 26). Because these patients were treated during the time period before the development of antiretroviral therapy, they comprised a control group of HIV-positive patients who had TB but did not receive antiretroviral therapy, and these patients were termed group 3. Sex and gender distribution in the three groups was as follows: group 1 (n = 33), 17 men and 16 women with a mean age of 39.9 years; group 2 (n = 31), 26 men and five women with a mean age of 43.3 years; and group 3 (n = 26), 19 men and seven women with a mean age of 34.6 years.
All chest radiographs obtained of patients in groups 1, 2, and 3 were collected. The radiographs were ordered chronologically and reviewed by agreement between two radiologists who were unaware of each patient's group affiliation. The radiographs were reviewed sequentially (i.e., time-lapse) without prior or simultaneous viewing of the whole series and without knowledge of the time interval between radiographs. The admission radiographs were evaluated for three categories of findings: pulmonary parenchymal opacities, intrathoracic adenopathy, and pleural effusions. Each subsequent radiograph was then compared with the previous one; in each of the three categories an assessment was made whether the findings were severely worsened, mildly to moderately worsened, unchanged, or improved. Severe worsening was defined as either the unequivocal appearance of new disease in a previously normal region of the chest or a doubling of the previous extent of disease (either in amount of opacified lung or in the size of enlarged nodes or pleural effusions). Mild to moderate worsening was defined as a visible increase in disease but qualitatively less than doubling. After all radiographs were reviewed, cases were categorized as to whether the series of radiographs showed: sequentially improving or unchanged findings; one or more occurrences of mild to moderate worsening; or one or more occurrences of severe worsening. For all occurrences of severe radiographic worsening, the clinical record was then consulted to determine whether there was concurrent clinical worsening as defined previously and whether a specific reason for radiographic or clinical worsening (or both) had been established. Alternate diagnoses to explain worsening were identified in two group 1 patients (pneumocystis pneumonia and bacterial sepsis) and one group 2 patient (nephrotic syndrome with pleural effusions); these three patients were excluded from subsequent analysis, leaving 31 patients in group 1 and 30 patients in group 2.
Statistical analysis was performed using the Fisher's exact test to compare the frequency of radiographic worsening among the groups. Among group 1 patients, CD4+ counts of patients who showed severe worsening and those who did not were compared using the Mann-Whitney test.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
Radiographic findings in the seven group 1 patients with severe worsening are listed in Table 3. Severely worsened pulmonary parenchymal disease was observed in four patients (Figs. 1A, Figs. 1B, Figs. 1C, Figs. 1D and 2A, 2B, 2C), hilar or mediastinal adenopathy in two patients (Fig. 3A, Fig. 3B), and pleural effusion in one patient. In four of the seven patients, the focus of radiographic worsening was in a previously normal region on the admission radiograph. Of the seven patients, five showed simultaneous mild to moderate worsening in a second category of findings. Radiographic worsening was observed between 2 and 32 days after initiation of antiretroviral therapy, whereas these seven patients had started inpatient TB therapy between 5 and 19 weeks before worsening (mean, 12 weeks). In six of the seven patients, worsening on the chest radiograph was concurrent with fever (n = 5), cervical lymphadenopathy (n = 2), ascites (n = 1), and cutaneous tuberculosis (n = 1). Baseline CD4+ cell counts before initiation of antiretroviral therapy were 47 ± 33 cells/mm3 (0.047 ± 0.033 x 109/l) in patients who subsequently showed severe radiographic worsening and 145 ± 196 cells/mm3 (0.145 ± 0.196 x 109/l) in other group 1 patients (differences not significant). There was no difference in the absolute increase in CD4+ count after treatment, with a mean CD4+ count increase of 38 cells/mm3 among worsening patients and 43 cells/mm3 among nonworsening ones. Four of the seven patients had PPD skin testing before beginning antiretroviral therapy; all were initially anergic. All four patients showed conversion of findings from the PPD skin test to positive between 4 and 20 weeks after antiretroviral therapy was started.
|
|
|
|
|
|
|
|
|
|
Among all group 1 patients with radiographic worsening (any degree), the mean time from beginning of worsening to beginning improvement was 7 weeks for pulmonary disease, 13 weeks for adenopathy, and 4 weeks for effusion. The times to improvement in patients in groups 2 and 3 were not significantly different. Of the seven group 1 patients with severe radiographic worsening, six patients showed radiographic improvement and one patient did not show radiographic improvement, although radiographs were only available for this patient up to 5 weeks after worsening was first identified (Table 3). All six patients in whom radiographic worsening was accompanied by clinical worsening showed clinical improvement, including the patient who failed to show radiographic improvement. In six of the seven patients with severe worsening, no changes were made to antituberculous therapy or antiretroviral therapy after clinical evaluation failed to show another reason for worsening. In one patient with severe worsening (Fig. 1), the possibility of drug reaction was considered and both antituberculous therapy and antiretroviral therapy were discontinued after superimposed infection had been excluded, after which radiographic and clinical improvement were seen. When this patient was subsequently restarted on antituberculous therapy and antiretroviral therapy, transient clinical worsening ensued but no radiographic worsening was noted.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A diagnosis of paradoxical response should not be made until noncompliance with therapy, drug resistance, superimposed disease processes, and drug reaction have been excluded as causes of apparent worsening. In our patients, noncompliance was not a consideration because all patients were hospitalized throughout the course of therapy. Resistance testing was performed monthly, and no patient developed drug resistance during this study. It is unlikely that minor variations in TB drug combinations among patients caused the differences among patient groups because treatment guidelines have not changed significantly over time [11, 12] and because worsening was more temporally related to antiretroviral therapy than to TB therapy. All patients with worsening were examined for the possibility of superimposed disease processes, which were identified in three patients. Finally, drug reaction may cause clinical and radiographic worsening in immunocompromised patients but was considered unlikely in our patients because the worsening resolved without changes or interruption of therapy for six of the seven patients; in the patient whose medications were discontinued (Fig. 1), the diffuse miliary pattern on the radiograph was not characteristic of drug reaction [14].
Clinical restoration of cell-mediated immunity after antiretroviral therapy in AIDS patients was first described by French et al. in 1992 [15]. In that study, 27 (42%) of 64 patients who had been anergic to tuberculin before AZT therapy subsequently developed a positive skin test response. Five patients also manifested an acute illness consisting of localized Mycobacterium avium-complex infection, lymphadenopathy, or fever (alone or in combination) after 1-2 weeks of therapy. In a more recent report, five latestage HIV-infected patients (CD4+ count, <<50 cells/mm3 [0.05 x 109/l]) who were started on antiretroviral therapy with the protease inhibitor indinavir developed fever, leukocytosis, and generalized lymphadenopathy within 3 weeks of commencing therapy [16]. Nodal biopsy results revealed granulomatous inflammation with M. avium-complex infection. The authors hypothesized that these patients had untreated subclinical M. avium-complex infection that was unmasked by the immunorestorative effects of antiretroviral therapy, resulting in an inflammatory lymphadenitis. Evidence for immune restoration included increases in CD4+ count, the presence of granulomatous features on biopsy specimens, and the vast preponderance of the "memory" CD4+ phenotype after treatment, suggesting an expansion of the CD4+ subset that had T-cell receptors to mycobacterial antigens resulting from the (previously subclinical) infection. All patients improved after initiation of antimycobacterial therapy. In our study, PPD reactivity was restored in patients with less than 200 CD4+ cells/mm3 (0.2 x 109/l), below which level most HIV patients remain anergic. This finding also suggests a selective restoration of immune function of the "memory" type. It has been shown in vitro that antiretroviral therapy selectively restores some but not all CD4+ T-cell surface marker phenotypes [17], which may help explain the lack of a large rise in the overall CD4+ count among our patients.
We recognize several limitations to our study. There were some differences in age and gender distribution among the three groups. This study occurred at a dedicated TB treatment facility at which most patients had been noncompliant with outpatient therapy, necessitating hospitalization. Thus, patients had received at least some antituberculous treatment at the time of admission, albeit suboptimal or intermittent. The fact that many of the patients (23%) had normal findings on chest radiographs at the time of admission could be related to prior therapy. Because of the high rate of noncompliance, antiretroviral therapy had generally not been started before admission as a result of concerns about development of HIV resistance. Patients who have not received prior antiretroviral therapy may be particularly at risk for paradoxical responses. Furthermore, most of our patients did not receive a protease inhibitor, which is the most highly active antiretroviral therapy; such combination therapies are becoming more frequently used, which may affect the frequency of paradoxical responses. Our use of a historical control group is limited by possible differences in the relative degree of immunosuppression in the two groups of AIDS patients (which could affect the relative incidence of paradoxical response) as well as in the degree of virulence of the TB organism. However, use of an HIV control group not receiving antiretroviral therapy is not feasible in the current era of AIDS therapy. Two aspects of radiographic evaluation in these patients also deserve mention. First, the differentiation between mild to moderate and severe worsening was qualitative; agreement between two radiologists was required but no prestudy training sessions were performed. On the other hand, there is not a commonly used or agreed on rating scale for chest radiographs of patients with TB, and our goal was to identify radiographic changes and not to quantitatively assess the amount of disease. Second, paradoxical responses did not always manifest as severe radiographic worsening. Five patients showed clinical signs or symptoms of paradoxical response, three of whom showed no radiographic worsening and two of whom showed only mild to moderate worsening. All five improved with continued therapy.
In conclusion, transient and occasionally severe chest radiographic worsening may be seen in AIDS patients undergoing antituberculous therapy subsequent to the introduction of antiretroviral therapy. Worsening may manifest as an increase in preexisting consolidation, lymphadenopathy, or effusion or may occur as new findings in previously normal areas. These episodes are similar in characteristics and timing to previously reported paradoxical responses in patients who begin antituberculous therapy. These phenomena may be a result of improving immunologic function, as suggested by the restoration of PPD reactivity in all previously anergic patients who showed severe worsening. After excluding superimposed infections, noncompliance with therapy, drug reaction, and drug resistance, patients exhibiting paradoxical responses should be maintained on combined therapy with consideration given to the addition of steroids to control local symptoms [10, 18].
Acknowledgments
We thank Arthur Pitchenik for valuable assistance with this project.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. W. Hull, P. Phillips, and J. S. G. Montaner Changing Global Epidemiology of Pulmonary Manifestations of HIV/AIDS Chest, December 1, 2008; 134(6): 1287 - 1298. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Masur Management of Patients with HIV in the Intensive Care Unit. Proceedings of the ATS, January 1, 2006; 3(1): 96 - 102. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N R Goldsack, S Allen, and M C I Lipman Adult respiratory distress syndrome as a severe immune reconstitution disease following the commencement of highly active antiretroviral therapy Sex Transm Inf, August 1, 2003; 79(4): 337 - 338. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. J. BURMAN and B. E. JONES Treatment of HIV-related Tuberculosis in the Era of Effective Antiretroviral Therapy Am. J. Respir. Crit. Care Med., July 1, 2001; 164(1): 7 - 12. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. B. Haramati, G. Lee, A. Singh, P. L. Molina, and C. S. White Newly Diagnosed Pulmonary Sarcoidosis in HIV-infected Patients Radiology, January 1, 2001; 218(1): 242 - 246. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
